Many people interested in human behavior do not feel the need for the standards of proof characteristic of an exact science; the uniformities in behavior are "obvious" without them. At the same time, they are reluctant to accept the conclusions toward which such proof inescapably points if they do not "sense" the uniformity themselves. But these idiosyncrasies are a costly luxury. We need not defend the methods of science in their application to behavior. The experimental and mathematical techniques used in discovering and expressing uniformities are the common property of science in general. Almost every discipline has contributed to this pool of resources, and all disciplines borrow from it. The advantages are well established.